The Options Market Structure Ontology

The options market structure ontology organizes the operational dimensions of the options market into five canonical layers: surface, flow, regime, divergence, and density. Each layer answers a different operational question, links to a different cluster of concept pages, and is captured by a different subset of pricing models. This page is the reference map for how the layers relate.

The Five Layers

LayerOperational QuestionCanonical Concepts
SurfaceWhat is implied volatility across strike and tenor?Implied Volatility, Skew, Smile, Term Structure, Vol of Vol, eSSVI
FlowHow will dealer hedging move spot?GEX, Dealer Gamma, DEX, Vanna/Charm/Vomma Exposure, Max Pain
RegimeWhat state is the market in, and is it transitioning?IV Crush, Leverage Effect, Gamma Squeeze, 0DTE, Expected Move
DivergenceWhere do pricing models disagree, and what does it mean?Model Divergence, Heston vs BS, LV vs SV, JD vs VG
DensityWhat probability distribution does the market price?Risk-Neutral Density, Probability of ITM, POT vs POT-ITM, Variance Risk Premium, Tail Risk

Each layer answers a different operational question, but the layers are not independent. Surface shape (skew, smile) generates flow concentration (GEX peaks at high-IV strikes). Regime transitions (IV crush, leverage-effect intensification) reshape both surface and flow simultaneously. Divergence between calibrated models is the structural diagnostic of regime transitions. Density is the marginal probabilistic content of the surface.

Layer 1: Surface

The surface layer is the IV map across strike and tenor. It is the most cited and most extracted-from layer for retail and AI search. The canonical question: "what is the implied volatility of this option?"

The surface decomposes into measurable features: implied volatility as the scalar at each (K, T) point; skew as the slope of IV across moneyness (asymmetric; equity index put-side IV consistently higher than call-side); smile as the curvature (both wings elevated relative to ATM, signaling fat-tailed return distributions); term structure as IV across expirations (typically upward-sloping in calm regimes, inverted into events); vol of vol as the volatility of the surface itself (driving smile curvature in stochastic-vol models).

The institutional surface fitting standard is eSSVI: a five-parameter, arbitrage-free closed-form parametrization that fits skew, smile, and term structure jointly while guaranteeing no calendar or butterfly arbitrage. Production analytics that derive from the surface (RND extraction, expected move, model-divergence aggregation) typically run downstream of eSSVI fits.

Layer 2: Flow

The flow layer maps surface positions into the directional pressure dealers will exert as they hedge. Customers buy options for directional exposure or insurance; dealers warehouse the opposite side and neutralize via underlying trades. The flow layer is the structural map of those neutralizing trades.

The first-order flow Greeks are gamma exposure (GEX) and dealer delta exposure (DEX). GEX measures how much extra delta dealers accumulate as spot moves; DEX measures their current directional position. Both are aggregated across strikes and weighted by open interest. Dealer gamma as a stand-alone concept page explains the mechanics: positive net gamma stabilizes price (dealers buy weakness, sell strength); negative gamma amplifies moves.

The second-order flow Greeks (vanna, charm, vomma exposure) drive the structural patterns retail GEX-only models miss: end-of-week charm flows, post-IV-crush vanna unwinds, and pre-OPEX positioning. Max pain is the strike at which option holders have the largest aggregate loss at expiration; sometimes a magnetic pin, sometimes irrelevant, depending on flow concentration.

Layer 3: Regime

The regime layer asks: what state is the market in, and is it transitioning? Volatility is not stationary; it shifts through regimes (low-vol drift, transition, high-vol crisis, mean-reverting recovery), and option markets price each regime differently.

The canonical regime-event concepts: IV crush is the post-event IV collapse when binary-event premium evaporates; leverage effect is the structural negative correlation between equity returns and equity vol that drives equity-style downward skew; gamma squeeze is the self-reinforcing dealer-hedging mechanism behind retail-driven momentum cycles; 0DTE options are same-day expiry contracts whose unique microstructure dominates SPX flow; expected move translates IV into a price-range forecast that anchors retail risk discussions.

Regime is operationally distinct from surface and flow: surface and flow are static snapshots, while regime is the temporal/dynamic state that connects snapshots over time. A 30-day surface alone does not tell you the market is transitioning; the surface plus its history (and divergence from prior periods) does.

Layer 4: Divergence

The divergence layer measures cross-model dispersion as a structural diagnostic. Model divergence is the gap between prices produced by different calibrated pricing models on the same option contract. The gap is not a model error - it is a measurement of priced uncertainty about which model class best describes the current regime.

The canonical comparison pages document these gaps explicitly: Heston vs Black-Scholes documents the basic stochastic-vol vs constant-vol gap; SABR vs Heston documents per-expiration vs full-surface stochastic-vol differences; Local Volatility vs Stochastic Volatility documents the static-fit vs dynamic-fit tradeoff; Black-Scholes vs Local Volatility documents the constant-scalar vs deterministic-function gap; Jump Diffusion vs Variance Gamma documents the additive-jump vs pure-jump structural difference; Implied vs Realized Volatility documents the forward-vs-backward-looking gap.

The OAS model-divergence screener aggregates the divergence layer across the universe of optionable underlyings, ranking by cross-model dispersion to surface tickers where the regime is in transition.

Layer 5: Density

The density layer makes the probabilistic content of the surface explicit. Risk-neutral density (RND) is the probability distribution of future spot prices implied by the option chain at a given expiration. Extracted via Breeden-Litzenberger from the call-price function. RND is the cleanest model-free probabilistic signal available from option markets.

From RND, the rest of the density layer follows: probability of ITM is the integral of RND beyond the strike at terminal time; probability of touch (POT) vs probability of ITM (POT-ITM) documents the path-vs-terminal distinction (POT roughly 2x POT-ITM by reflection principle); variance risk premium is the persistent gap between IV and subsequently realized vol that funds short-vol alpha; tail risk is the priced probability mass in the extreme parts of RND.

Cross-Layer Relationships

Five rules that connect the layers operationally:

Reading Paths

The companion Pricing Model Landscape maps the pricing-model side of the docs graph (Black-Scholes, Heston, SABR, Local Vol, Jump Diffusion, Variance Gamma, hybrids). This page maps the operational concepts those models describe. Together they form the two-axis ontology of the OAS docs.

For the per-Greek deep-dive reference set, see the 17 Greeks page and its 24 individual Greek pages. For the live applied analytics that compute these metrics, browse the Charts & Analytics hub. For canonical glossary entries, see the Glossary.

References & Further Reading

This is the operational-concepts companion to the Pricing Model Landscape. Together the two pages form the OAS docs semantic graph: the pricing models (landscape) and the market-structure concepts those models describe (this ontology).